Use snp file to find the input impedance in ads

[chenyiting]

I'm trying to design a Wilkinson power divider with HFSS, but I need to add a lambda/4 line between sqrt(2)Zo part and the Zo part as the pic shows for impedance matching , so I want to know the input impedance there. Can I do this with ADS by importing S parameter matrix to a SnP component in schematic?

 

[FvM]

Not sure which lambda/4 line you refer to. By design, Wilkinson divider consists of two Z0*sqrt(2) lines and no additional matching elements. https://en.m.wikipedia.org/wiki/Wilkinson_power_divider

 

[chenyiting]

Not sure which lambda/4 line you refer to. By design, Wilkinson divider consists of two Z0*sqrt(2) lines and no additional matching elements. https://en.m.wikipedia.org/wiki/Wilkinson_power_divider

Yes, but however between the 50 ohm and 70.7 ohm line still have some mismatch, I want to add another lambda/4 line to make it matches better. The point here is to use s parameter from HFSS and the length of 50 ohm line in ads to find the value of input impedance, but I have no clue how to do it.

 

[FvM]

Doesn't make sense to me. If you observe impedance mismatch, the actual length or impedance of the 70.7 ohm lines isn't exactly right and should be tuned. Also the geometry of the junction point can be adjusted.

 

[chenyiting]

Doesn't make sense to me. If you observe impedance mismatch, the actual length or impedance of the 70.7 ohm lines isn't exactly right and should be tuned. Also the geometry of the junction point can be adjusted.

Sorry for being misleading, this Wilkinson power divider is part of my graduate school training program to let me get use to the software. I just want to know if today I have S parameter and the length of the feeding line, can I find the input impedance at the end of the feeding line with ADS. The impedance matching part is actually doesn't matter.

 

[volker@muehlhaus]

One issue unrelated to your question: Have you included vias in your layout to connect the side grounds to bottom ground?

Regarding de-embedding the feedline, you might get the line impedance and effective epsilon from HFSS port information, then you can remove it by cascading a negative line length in ADS.
Or you can simulate that feedline as a separate model for de-embedding purposes, using the line S2P data. Search for De-embed2 element in ADS help:
Home > Components > Components-Analog RF > Introduction to Circuit Components > Data File Components